"> Specifically I'm not quite sure I know what you mean by "calculate

the

average contracts for ... during the yr" and "... do not have the same

number of months"."

First I would like to thank you for responding to my question and

appologize for my lack of clarity. I have been using DB2 on a

mainframe for about 3 years now at work.(I am an actuary for an

insurance company). Recently I had a OBDC driver installed on my PC. I

have never used ACCESS before so I thought this might be a good

oportunity to learn it. Lets say that I need to count the average

number of monthly contracts for Wal-Mart("contracts" are the insured

individuals or families.) Wal-Mart has been with my company for the

entire year and this is October so I have 10 months of contracts to

count/average. I add up the number of contracts for each month, which

fluctuates, and divide by the number of months to get an average. If

all the 100 or so groups had a full 10 months of contracts it would be

simple. I would just divide the sum for each group by 10. The first

part of the problem arises when a group like United Air or Food Lion

comes on board in April or May. In this case I do not have 10 months

of data so I cant just divide by 10. I have to divide by 5 or 4

months respectively. The second part of the problem is that each group

can have thousands of entries on the DB2 table that I am hitting. To

count the months I took advantage of the fact that every month has a

1st day. Jan 1st, Feb 1st....Oct 1st. I count the number of distinct

1st days of each month and divide it into the total number of

contracts. This simple technique gives me accurate results when using

DB2. More about the report: The report is simple in description but

very difficult in practice(the story of my life) I am looking at all

the monies we charge a group(admin fees) and all the monies we are

paying out(primarily network access fees) on a per state basis. So I

really need to know the average number of contracts in each

state/country(which I break out later). The contract counts are a

VERY small part of the report. The DB2 proc that I hope to replace

with access runs 17 queries pulling data from 9 different DB2 tables.

When all is said and done the final report is about 250 pages. So

what I need to know is how to calculate the average number of

contracts per month over the number of months in the current year that

each particular group has been a client of my company.(Wow thats a

mouth full) One more thing I know that my method is probably not the

best way of accomplishing the task at hand but I am "self taught" so

have had no formal SQL training thus a limited tool box. In college I

was a physics/math major programming primarily with fortran and LaTex.

Thanks